Some existing location services rely on crowd-sourced data to deliver location information to requesting computing devices such as mobile telephones. The existing systems, however, assume that all the beacons are stationary. In practice, some of the beacons may move or be moving, which may result in multiple probable locations for the beacon. Some existing location services attempt to identify the multiple probable locations for the beacon by performing a clustering analysis on the crowd-sourced data. The clustering analyses, however, become very complicated (e.g., time consuming and computationally intensive) for beacons that have moved more than once.